Five questions about data visualization in early case assessment

Five questions about the role of data visualization in early case assessment

What is the difference between data visualization and document review tools?

Both are text analytics tools that use keywords to create, search, and classify a corpus of documents. Some document review tools are used primarily to prioritize and review individual documents to determine their responsiveness after the corpus of documents has been culled and classified, typically with a separate tool. A visualization tool, such as Brainspace Discovery5™, does not perform individual document reviews, but it is an effective culling tool that efficiently produces visual representations of document categories for prioritization purposes.

But what does a data visualization tool do that is so different?

For prioritization and review purposes, a document review tool typically produces a report that lists document categories and words from those categories associated with each document. The report resembles a spreadsheet, and although it provides clustering, it’s often not very intuitive or easy to review, and it can take weeks for power users to prioritize categories and individual documents, depending on the scope of documents to be reviewed.

A visualization tool, when linked with a document review tool, takes the corpus of documents and produces visual representations of the data. Brainspace Discovery5, for example, produces a “cluster view”—an interactive wheel graphic that shows “search hit” categories and allows power users to drill down to individual documents within those categories. Power users can actually open and review the document textual contents, though through the visual representation of the categories, they can much more quickly gain a sense of which documents should be prioritized for review. In addition, for communication analysis purposes, power users can see separate visualizations depicting relationships between authors of e-mails—a capability that helps significantly speed the e-mail prioritization or identify e-mail patterns not typically obvious.

What benefits do data visualization tools offer?

The biggest benefit is the visual representation of the data. Given the volume of data most companies possess today, it can be overwhelming. A visualization tool like Brainspace Discovery5 can help users wrap their arms around that data, understand it, categorize it, and decide how it tells the story about the case.

Speed and cost-efficiency are other key benefits. With standard document review tools, attorneys can spend days, if not weeks, going back and forth testing search terms based on a list of numbers. Visualization tools enable power users, in just a matter of hours in many cases, to create an index of search terms, visually review the documents containing those terms, and determine whether those terms are hitting the mark. If not, revising the terms is fast and easy, with near instantaneous visual results. Quick identification of document clusters that don’t need to be review saves significant time, which helps with controlling eDiscovery spending.

Yet another benefit is the ability to incorporate opposing counsel’s data into the visualization tool and quickly assess what their data looks like. We recently did this for a client involved in litigation where there were nearly 30 other parties. We had data productions from those other parties, so we created a Brainspace Discovery5 index that allowed our client to rapidly search and prioritize all of that other data.

Finally, a visualization tool allows power users to see a visual representation of relationships between subject parties and the people they communicate with via e-mail. You can quickly run specific searches on individuals—or e-mail domain names—and produce visual maps of all the other people that person or domain name has communicated within that data set.

How does data visualization work in a typical early case assessment scenario?

After initial processing and de-duplication of the data set that the client brings to us, we host that data in a staging environment. For example, in our Electronic Discovery Solutions Center, we often suggest that users start with a visualization tool—Brainspace’s Discovery5 is one of the tools we use—to speed the process of identifying documents the user would like to push to review. We typically train users on visualization tools so they keep costs as low as possible. In a short time, they can identify and refine their search terms, create “collections” or clusters of documents, analyze and “mass tag” documents in those collections to determine which ones should be reviewed for responsiveness, and then export the collections back over for the review process.

Can data visualization be used for other investigative purposes?

Recently, we have leveraged Brainspace Discovery5 to help users categorize large groups of contracts that were not organized in any fashion. By importing the contracts into Brainspace Discovery5, we can use the cluster view to automatically categorize the contracts provided in the data set. Then our contract review specialists can verify whether those categories are responsive to the client’s objective for organizing the data into defined data sets. We believe there will be other similar uses for visualization elsewhere in client organizations.

Our take

The eDiscovery challenges continue to grow as data volumes and sources continue to expand. With the “big data” surrounding organizations today, legal departments and their external counsel struggle to quickly get a handle on the data needed for a given investigation or litigation, and it is almost always difficult to review that data in a strategic manner to make decisions during early case assessment. Visualization tools represent a breakthrough in terms of both time- and cost-savings to speed the discovery process and assist counsel and clients in making better-informed decisions.